mradermacher's picture
auto-patch README.md
3789fca verified
metadata
base_model: fblgit/una-xaberius-34b-v1beta
datasets:
  - fblgit/tree-of-knowledge
  - garage-bAInd/Open-Platypus
  - allenai/ultrafeedback_binarized_cleaned
  - Open-Orca/OpenOrca
language:
  - en
library_name: transformers
license: cc-by-nc-nd-4.0
quantized_by: mradermacher
tags:
  - UNA
  - juanako
  - cybertron
  - xaberius

About

weighted/imatrix quants of https://huggingface.co/fblgit/una-xaberius-34b-v1beta

static quants are available at https://huggingface.co/mradermacher/una-xaberius-34b-v1beta-GGUF

Usage

If you are unsure how to use GGUF files, refer to one of TheBloke's READMEs for more details, including on how to concatenate multi-part files.

Provided Quants

(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)

Link Type Size/GB Notes
GGUF i1-IQ1_S 7.6 for the desperate
GGUF i1-IQ1_M 8.3 mostly desperate
GGUF i1-IQ2_XXS 9.4
GGUF i1-IQ2_XS 10.4
GGUF i1-IQ2_S 11.0
GGUF i1-IQ2_M 11.9
GGUF i1-Q2_K 12.9 IQ3_XXS probably better
GGUF i1-IQ3_XXS 13.4 lower quality
GGUF i1-IQ3_XS 14.3
GGUF i1-Q3_K_S 15.1 IQ3_XS probably better
GGUF i1-IQ3_S 15.1 beats Q3_K*
GGUF i1-IQ3_M 15.7
GGUF i1-Q3_K_M 16.8 IQ3_S probably better
GGUF i1-Q3_K_L 18.2 IQ3_M probably better
GGUF i1-IQ4_XS 18.6
GGUF i1-Q4_0 19.6 fast, low quality
GGUF i1-Q4_K_S 19.7 optimal size/speed/quality
GGUF i1-Q4_K_M 20.8 fast, recommended
GGUF i1-Q5_K_S 23.8
GGUF i1-Q5_K_M 24.4
GGUF i1-Q6_K 28.3 practically like static Q6_K

Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better):

image.png

And here are Artefact2's thoughts on the matter: https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9

FAQ / Model Request

See https://huggingface.co/mradermacher/model_requests for some answers to questions you might have and/or if you want some other model quantized.

Thanks

I thank my company, nethype GmbH, for letting me use its servers and providing upgrades to my workstation to enable this work in my free time.